跟读练习: Inside a warehouse run by AI robots - 通过YouTube学习英语口语
C1
We're enabling the next generation of Amazon to really grow
149 句
如果句子过短或过长，请点击 Edit 进行调整。
1
We're enabling the next generation of Amazon to really grow
2
and realize their full potential and driving the economy forward because this is the future of logistics.
3
I'm Andre, Chief Growth Officer at BrightPick,
4
and we're developing the next generation of warehouse robotics.
5
Our mission is really to enable any warehouse,
6
large or small, to fully automate their order fulfillment in a matter of weeks.
7
So right now we're standing in one of our installations
8
where we pretty much automated the entire fulfillment process using close to 50 BrightPick Autopicker robots.
9
I'm going to show you around so you can see how it looks.
10
So our flagship robot is called BrightPick Autopicker
11
and it's actually the only mobile robot in the world that not only transports items but can also robotically pick them.
12
And the big advantage with that is it actually does the pick directly in and out.
13
And we actually took a lot of inspiration from the way humans do picking because humans,
14
they go directly from one pick to the next and that is actually the most efficient way to pick orders.
15
Basically, what we did is you can think of BrightPick Autopicker
16
as almost like a human with a cart moving from one location in the warehouse to the next
17
and doing the picks as it goes on.
18
In fact, we're getting pretty close to human level performance and at a fraction of the cost.
19
So the way the robotic picking process works is the robots,
20
they retrieve a storage tote from the shelving,
21
they scan the storage tote,
22
create a 3D model of the tote.
23
Then we have an AI engine which tells the robot what item to pick based on probability of success,
24
and then the robot picks that item from the storage toad into the order toad.
25
The thing with robotic picking is that inevitably the robots will not be able to pick some items some of the time.
26
I think anyone who tells you they can pick all items 100% of the time are lying, frankly.
27
It's a big challenge in robotic picking in terms of what do you do as a fallback in that scenario.
28
We're actually not entirely relying on robotic picking here.
29
We incorporated a couple of goods-to-person stations into this installation,
30
both as a fallback in case the robots can't pick the items,
31
but also for items that aren't ideally suited for robotic picking.
32
So let's say items that are heavier or that have a changing center of gravity,
33
like liquids, for example.
34
So this is our goods to person station right here.
35
It actually works on the same principle as the robotic picking.
36
So each auto picker brings both the storage tote and the order tote to a human operator.
37
There's a tablet that tells him exactly how many items and
38
which items he needs to pick and then the operator picks it from one tote to the other tote.
39
So it's super simple, pretty ergonomic.
40
The robots they raise both totes to waist height.
41
They actually have four different sides.
42
So what that enables us to do is to feed the human operator from four sides at the same time.
43
So during peak hours when it's really busy,
44
there's basically four robots at the station at any given point in time,
45
which means there's like zero idle time in between fixed.
46
A really important part of our philosophy is product simplicity and cost.
47
You can see like we use a pretty simple SCAR arm on the robots.
48
It's two axes, the end effector is pretty flexible
49
and we could have of horse stuck like an expensive six-axis arm on the robot,
50
but that would have basically doubled the cost for an extra 10% of performance.
51
There's no point in doing that
52
when basically what we do in those 10% of the edge cases is we take those items to a human operator,
53
which you would anyways need because even if you have the most expensive robotic picker in the world,
54
there's still going to be some items when the robot can't pick them.
55
So you anyways have to solve for that.
56
And I think that's I think that's a really important part to understand about the whole BrightBix system is
57
it's designed with simplicity in mind.
58
So there's a lot of super complicated features,
59
AI, you know, the software orchestration behind it.
60
But the actual process of working with the robots and installing it in the warehouse,
61
it's super simple, right?
62
You see I'm like walking around with the robots.
63
Of course, usually we all want people to do that because it slows down the system.
64
But it's perfectly safe, you know,
65
you don't need any safety fencing.
66
There's no QR codes on the floor for navigation.
67
We use LiDARs and AI.
68
The shelving, it's like very cheap.
69
It's basically the same shelving you can buy in Home Depot,
70
which also means like you can install in a matter of weeks and even move the shelving from one warehouse to another,
71
which, you know, it's pretty important.
72
You don't want to build like a monument in your warehouse that you're stuck with for the next 10 years.
73
That's actually the reason we purposefully don't use any conveyors because,
74
you know, once you build a conveyor good luck moving that.
75
That's basically why we developed the system is to enable any retailer to automate their fulfillment in a matter of weeks.
76
This is really like a decade of development behind this.
77
I mean first we developed the 3D scanner which is actually from our sister company Photo Neo.
78
We then developed the AI picking engine.
79
I think we've done close to a billion picks by now to train our AI.
80
We then developed a mobile platform,
81
all the software orchestration that goes with that.
82
We actually had a goods-to-person solution before this,
83
but then ultimately we brought it all together inside the Autopicker.
84
And now this robot can do everything you need for order fulfilling, right?
85
Like robotic picking, goods-to-person picking,
86
pallet picking, order buffering, consolidation, dispatch, replenishment.
87
When we think of BrightPick Autopicker as basically a humanoid in all but name,
88
If you think about what a human picker does in a warehouse,
89
it's basically the same, except auto pickers are able to do it much faster, more efficient.
90
It's literally a tenth of the cost of a humanoid.
91
We don't believe in technology for technology's sake.
92
Why do you need a human form factor when you can basically achieve human level performance at a fraction of the cost?
93
So a common misconception that people have about robotics is that we're taking away people's jobs,
94
but the reality is actually completely opposite because pretty much every customer that buys our solution,
95
they're dealing with massive labor shortages,
96
whether it's in logistics or in manufacturing,
97
and there's simply not enough people to meet demand because we have a shrinking labor force.
98
These are physically tough jobs.
99
Imagine walking two marathons a week just to pick orders,
100
and there's not enough people to fill the demand, basically.
101
For example, this customer right here,
102
they had to actively turn off their marketing spend because they didn't have enough people in their warehouse to pick orders.
103
And now after they installed our solution,
104
they're able to grow it 100%.
105
It was basically like three people servicing our entire system.
106
Whereas before they would have needed 20, 30 people here.
107
Another feature that's unique to BrightPick is what we call pallet picking.
108
And this is actually, imagine you have items that are fast movers.
109
So let's say you're putting in samples or you're basically selling products that go into every order or every second order.
110
It doesn't really make sense to put those items into the system
111
because you're going to be replenishing them almost as fast as you're taking them out.
112
So what we developed, what we call these pallet picking positions,
113
where basically human operators are able to pick directly from pallets or from cases,
114
which you see stored behind them,
115
and place them onto the order tote at the robot.
116
So that basically makes the replenishment process super easy,
117
where all you do is you stick a pallet there or you throw a case there,
118
and then the robot pulls up,
119
tells the human operator on the tablet what item to pick,
120
and the human can pick that item and place it onto the robot.
121
So you really get that flexibility of robotic picking,
122
GTP picking, and pallet picking,
123
all as part of one system.
124
You really don't need any other picking.
125
So this right here is the buffer area,
126
and actually this is a 24-7 operation,
127
so the robots are working even at night when there's practically no one in the warehouse.
128
and what they're doing is after they pick the orders,
129
they buffer them in this area.
130
And then once the morning shift comes in,
131
they take them out, deliver them to pack out to be packed and shipped.
132
Another use case is imagine you have an order that consists of 10 picks,
133
but only nine of them are robotically pickable.
134
So the robots are gonna pick those nine items overnight,
135
buffer them, and then when the morning shift comes in,
136
it's gonna do that last 10th pick using the goods to person or pallet station.
137
And basically then the order is complete.
138
So because we have such a high level of automation,
139
we're able to easily operate at night because you practically don't need anyone to run the system.
140
We're now in a hyper-growth mode.
141
We're really focused on ramping up our manufacturing because we already get more demand than we can meet for our solution.
142
So now it's really about laying down the groundwork to make
143
sure we can manufacture as many of these robots as possible and really ensure that they're as reliable as possible.
144
We understand that these are mission critical processes,
145
so we make sure that each robot can really operate five,
146
10 years without breaking down.
147
This is one of many installations that we've already done.
148
We have many more in the pipeline,
149
so right now it's rolling up our sleeves and getting to work.
下载应用
AI 为你说出的每个句子打分
TRENDING
热门
背景与上下文
在这段视频中，安德烈（Andre）介绍了全新的仓库自动化解决方案，讲述了未来物流的发展。作为BrightPick的首席增长官，他分享了有关智能机器人如何加速订单履行的理念，并强调了将人工智能融入仓储管理的必要性。通过对机器人的介绍，我们了解到这是一项能够促进经济发展的技术创新，尤其是在电商领域。此类技术的推广将使每个仓库都能在短时间内实现自动化，极大提高工作效率。
日常交流中的五个常用短语
- 我们正在推动经济发展 (We are driving the economy forward)
- 让任何仓库完全自动化 (Fully automate any warehouse)
- 机器人进行拣选 (Robots pick the items)
- 人类拣选的方式 (The way humans do picking)
- 货物到人工作站 (Goods-to-person station)
逐步影子跟读指南
为了提高您在观看此视频时的英语水平，您可以依照以下步骤进行影子跟读练习， 提高英语发音 和流利度：
- 观看视频片段：专心观看视频的前1-2分钟，注意讲者的发音、语调和节奏。
- 逐句模仿：暂停视频，逐句重复讲者的内容，尝试模仿他们的语音语调，特别是其中的关键短语。
- 关键词记忆：结合视频中的日常短语，进行反复训练，以便在日常交流中自如使用。
- 反复练习：反复进行影子跟读，直到您能够在没有停顿的情况下流畅输出。
- 自我录音：录下您自己的声音，然后与视频中的原声进行对比，注意发音和节奏的差异。
通过这样的影子跟读练习，您能有效提升口语能力，尤其在进行雅思口语练习时，将更加得心应手。同时，建议结合 看YouTube学英语 的视频，以丰富您的学习方式。
什么是跟读法？
跟读法 (Shadowing) 是一种有科学依据的语言学习技巧，最初开发用于专业口译员的培训，并由多语言者Alexander Arguelles博士普及。这个方法简单而强大：您在听英语母语原声的同时立即大声重复——就像是一个延迟1-2秒紧跟说话者的影子。与被动听力或语法练习不同，跟读法强迫您的大脑和口腔肌肉同时处理并模仿真实的讲话模式。研究表明它能显着提高发音准确性，语调，节奏，连读，听力理解和口语流利度——使其成为雅思口语备考和真实英语交流最有效的方法之一。